Recent work in our lab suggests that, during adolescence, asynchronous patterns of development across different brain systems produce a period of increased sensitivity to affective stimuli and diminished capacity for cognitive control. Especially in affectively salient peer contexts, this regulatory imbalance may contribute to adolescents' increased proclivity to engage in risky behaviors (e.g., substance use, unprotected sexual activity, and reckless driving) that significantly threaten their health and well-being. In separate work, we and others have demonstrated the potential to improve adults' cognitive control through targeted working memory (WM) training. In the present proposal, we aim to bridge these two lines of research by using behavioral and neuroimaging methods to test a novel application of WM training to promote adolescents' cognitive control capabilities, and thus reduce risk taking. We hypothesize that adolescents' control over impulsive and automatic behaviors, especially in affectively salient contexts, can be improved through a targeted program of adaptive complex WM training. We further hypothesize that such improved cognitive control will engender trained adolescent participants with enhanced deliberative decision making skills, and hence reduce their proclivity toward risk taking. Using fMRI, we will characterize the mechanisms that underlie these behavioral changes. The PI and co-PI have a recent history of successful collaboration, and combined expertise in the cognitive neuroscience of cognitive control and the psychosocial development of adolescents and young adults. The proposed studies have the potential to greatly inform theoretical models of adolescent behavior, to yield fundamental insights into the processes that affect adolescent decision-making, to determine whether the maturational timetable of cognitive control can be accelerated through training, and to yield a low-cost intervention for reducing adolescent risk behavior.